It is becoming feasible to generate massive quantities of DNA sequence data for disease association studies. This presents both challenges and opportunities for human genetics. Perhaps most importantly, with large scale resequencing data, it will be possible to start identifying genes at which rare variants contribute to disease susceptibility. Here we propose to create a number of the analytical tools that will be needed for analyzing and interpreting the forthcoming data. Our first two Aims focus on using some of the first genome-wide resequencing data to better annotate noncoding sites that are likely to be functional. Our third Aim develops statistical methods for analyzing data that emerge from disease association studies to identify genes with rare variants that contribute to disease. The statistical methods will use the annotation approaches that we will develop in the first two Aims to prioritize variants according to the likelihood that they might have biological function. Using, in part, our improved functional annotation of potentially functional sites, we will also develop new statistical methods to identify genes that contribute to disease phenotypes through the action of many rare variants. [unreadable] [unreadable] [unreadable]